This archive contains all scripts and data necessary to generate results in both the main paper and the online appendix. Scripts used to generate tables and figures from the main text have a suffix `main_`. Scripts used to generate tables and figures from the appendix have a suffix `a_`. All scripts clearly indicate the specific tables/figures they generate in the title. To run these scripts using the given file structure, create a folder called "ngo_afe_wp, download the R project included in this archive, and create subfolders called `data`, `tables`, and `plots`. Download all data to the `data` folder. At the time of publication, the stargazer::stargazer() function used in many of these scripts to generate tables will generate an error when using an R version 4.2 or later. This error results from the object names having more characters that the function is able to accept. To generate the tables in R 4.2 or later, rename each object to something shorter.

# Main Datasets:
dat_unrestricted.rds
* Unrestricted sample
dat_restricted.rds
* Restricted sample 

# Village-level data from Bjorkman 2018
* All datasets with prefix "AEJ" are from the replication files for Bjorkman (2018). These files provide baseline and endline data for the original RCT. See referenes in main paper for additional information.

# Main dataset with Credit Attribution variables not imputed so that we know how many DK/R answers there were (only for plotting)
dk_responses.csv

# Village-level data collected from VHTs and CHPs
village_d811a.rds
* treated control 
village_d811b.rds
* pure control

# Respondent participation in recent elections
election_panel.csv


